Algorithms in Multi-Stakeholder Interactions: A Case Study of the Home Valuation Tool Zestimate

The development and use of digital platforms to facilitate economic transactions are on the rise. Zillow, an online platform, and marketplace for real estate, offers its users an opportunity to sell, buy, rent, and finance on- and off-market properties. This study focuses on Zillow's algorithmic tool Zestimate, which estimates home values based on multiple data points and has become a popular tool among home buyers and sellers. In this case study, we explore real estate agents’ and home buyers’ and sellers’ perceptions and experiences of Zestimate. Preliminary findings provide insight into how each party may interact with Zestimate and how multi-stakeholder interactions are reconfigured by the algorithm. Adopting an ecosystemic perspective, we explore the algorithmic systems’ integration and complexity in its social context beyond singular human-AI interactions. Based on these findings, we discuss the implications for human-centered and socially sustainable design practices when designing algorithmic systems in a multi-stakeholder context.